prompt-architect

Pass

Audited by Gen Agent Trust Hub on Feb 17, 2026

Risk Level: SAFEPROMPT_INJECTION
Full Analysis
  • PROMPT_INJECTION (LOW): The skill possesses a surface for indirect prompt injection (Category 8). Evidence Chain: 1. Ingestion points: Processes free-text user tasks, JSON output from the clarify-spec skill, and local files (CLAUDE.md, AGENTS.md). 2. Boundary markers: Employs XML tags like <output_structure> and Markdown headers to segment instructions from data. 3. Capability inventory: Limited to reading local documentation and code for context; no network or file-write capabilities are present. 4. Sanitization: Relies on structural separation; no explicit escaping of untrusted input.
  • DATA_EXPOSURE (SAFE): Access is limited to project context files (CLAUDE.md, AGENTS.md); no sensitive credential paths are targeted.
  • REMOTE_CODE_EXECUTION (SAFE): No execution of remote scripts or unverified package installation detected.
Audit Metadata
Risk Level
SAFE
Analyzed
Feb 17, 2026, 06:28 PM